Mule deer habitat use in the Bennett Hills, Idaho
by Clint Jason Gray
A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Fish
and Wildlife Management
Montana State University
© Copyright by Clint Jason Gray (1995)
Abstract:
The factors governing mule deer (Odocoileus hemionus) habitat use patterns in south-central Idaho are
poorly understood. Therefore, 16 radio-collared does were tracked from the ground in 1993, and 9 of
the same does were also tracked in 1994 to gather data on habitat use and to document the effects of the
primary land use in the area, cattle grazing, on mule deer. The field season of 1993 was cooler and
wetter than the 1994 season. Average home range size during the hot and dry field season was
significantly smaller than the average home range size during the cool, wet season (P<0.01). Deer
showed a strong preference for the Mountain Brush cover type for all activities, Aspen and Riparian
cover types for bedding, and Sage/Bitterbrush for feeding. Cooler northeastern and eastern aspects
were preferred by deer and southwestern and flat aspects were avoided. Overall, deer appeared to
prefer steeper than average areas, but within Mountain Brush they chose flatter sites. Areas near water
were preferred and stock ponds may have made bedding areas that would have been too far from water
otherwise available for use. Deer preferred microsites that did not have fresh sign of cattle. Habitat use
appeared to overlap more between deer and cattle during the dry year and the latter part of the
summer/fall period. A more detailed analysis of the effects of cattle on mule deer was not possible due
to the high density of different landowners in the area. Hunting season did not affect bed concealment
or flushing distance of deer but may have caused movements out of established home ranges and
premature migration to the winter range. Deer chose shaded areas for bedding and preferred clumps of
Salix sp. shrubs within the Mountain Brush cover type. Microhistological fecal analysis documented
the transition from forbs to shrubs in late summer. A model built with multiple logistic regression
identified overhead cover as the most important factor for determining sites likely to be used by deer.
This study documented the importance of the Mountain Brush cover type to mule deer in sage steppe
environments. It also identified the potential of hunting pressure to trigger premature migrations. MULE DEER HABITAT USE IN THE
BENNETT HILLS, IDAHO
by
Clint Jason Gray
A thesis submitted in partial fulfillment
of the requirements for the degree
of
Master of Science
in
Fish and Wildlife Management
MONTANA STATE UNIVERSITY
Bozeman, Montana
June 1995
N^ig
G=-T6IlS
APPROVAL
of a thesis submitted by
Clint Jason Gray
This thesis has been read by each member of the thesis committee and has been
found to be satisfactory regarding content, English usage, format, citations, bibliographic
style, and consistency, and is ready for submission to the College of Graduate Studies.
/3
Date
TU 7 h ’j'T
V
Chairperson, Graduate Comijnjtfee
Approved for the Major Department
12 J u l y 1995
Date
Head, Major Department
Approved for the College of Graduate Studies
Date
Graduate Dean
STATEMENT OF PERMISSION TO USE
,
In presenting this thesis in partial fulfillment of the requirements for a master's degree
at Montana State University, I agree that the Library shall make it available to borrowers
under rules of the Library.
,
IfI have indicated my intention to copyright this thesis by including a copyright notice
page, copying is allowable only for scholarly purposes, consistent with "fair use" as
prescribed in the U.S. Copyright Law. Requests for permission for extended quotation from
or reproduction of this thesis in whole or in parts may be granted only by the copyright
holder.
Signature
Date___
ACKNOWLEDGEMENTS
I would like to thank Dr. Lynn Irby for supervising all phases of this study from
proposal to thesis. His assistance was invaluable. I also appreciate Drs. Harold Picton,
Jay Rotella and James Unsworth for critical review of the manuscript, and Dr. Thomas
McMahon for serving on my committee. The Idaho Department of Fish and Game
funded the study. Dr. James Unsworth efficiently provided equipment and field support.
A special thanks to Bruce Palmer, Craig Kvalle and the rest of the crew in Region 4 for
providing deluxe accommodations and scrounging needed equipment at the last minute.
My greatest debt of gratitude goes to my wife. Cherry, for her assistance in the field and
at the computer. Her patience and love in the face of multiple moves, poor living
conditions, desolation and poverty were greatly appreciated.
V
TABLE OF CONTENTS
Page
APPROVAL............................................................................................
ii
STATEMENT OF PERMISSION TO USE................
iii
ACKNOWLEDGEMENTS........................................; ........... ..............................
iv
TABLE OF CONTENTS ........................................................................................
LIST OF TABLES................................................................................................
LIST OF FIGURES ................................................................................................
v
viii
x
ABSTRACT..........................................................................................................
xii
INTRODUCTION................................................................................................
!
STUDY AREA........................................................................................................
5
METHODS .............................. <........................................................................
8
Capturing and Radio-collaring Mule D eer................................................
Deer Relocations ......................................
Random Locations.............................................................
Home Range Analysis ...........................
Statistical Analyses................................
Univariate Analyses ......................
Multivariate Analysis ................
RESULTS ..........................................................................................
Climatic Differences Between Y ears........................................................
Home Range Analysis ..............................................................................
8
9
12
13
13
13
14
16
16
19
Vl
TABLE OF CONTENTS - continued
Page
Cover Type Use ........................................................................................ 23
Overall Cover Type U s e ..............................................: ............... 23
Activity and Cover Use ................................................................ 24
Variability Between Y ears.........................
25
Aspect U s e ............................
25
Overall Aspect U se ........................................................................ 25
Variability Between Y ears.............................................. ! ........... 27
Relation to Mountain B ru sh .......................................................... 28
Topography U s e ........................................................................................ 29
Overall Topography Use ............................................
^ /2 9
Variability Between Y ears....................................................... 30
Horizontal Configuration Use .................................................................. 30
Overall Horizontal Configuration ................................................ 30
VariabilityBetween Y ears............................................................ 31
Elevation and Slope Use .......................................................................... 32
Distance to R oads...................................................................................... 33
Overall Distance to R o a d s............................................................ 33
Before and During Hunting S easo n .............................................. 33
Distance to W ater...................................................................................... 34
Overall Distance to Water ............................................................ 34
Variability Between Y ears............................................................ 34
Variability Between Seasons...........................................
35
Distance to Livestock................................................................................ 35
Overall Distance to Livestock ...................................................... 35
Variability Between Y ears............................................................ 35
Variability Between Seasons .................
35
Concealment and Distance When Flushed................................................ 36
Deer Bed Shading........................
36
Food Habits..........................
37
Multivariate Habitat Analysis ..............................
38
DISCUSSION ........................................................................................................ 40
Home Range ....................................................
Cover Type Use .................................. *....................................................
Aspect, Topography, and Horizontal Configuration ................
Elevation and S lo p e ..................................................................................
Roads, Water and Livestock .......................................................................
40
41
44
46
47
vii
TABLE OF CONTENTS - continued
' >
Page
Concealment and Flushing Distance ........................................................
Deer Bed Shading......................................................................................
Food Habits................................................................................................
Multivariate Analysis................................................................................
50
50
51
51
CONCLUSIONS .................................................................................................... 53
REFERENCES C IT E D .................................................................................
55
APPENDIX - FREQUENCY, MEDIAN, AND MEAN COVER CLASS
OF VEGETATION IN COVER TY PES................................................................ 60
I
viii
LIST OF TABLES
Table
1. Phenological stages and codes used for vegetation at Bennett Mountain . . .
Page
12
2. Description of variables used in the multiple logistic regression.................. 15
3. Numbers of locations and home range sizes for all deer tracked in the
Bennett Hills, 1993 and 1994 ........................................................................ 20
4. Distances, in meters, to primary and secondary roads from random and
deer locations, overall, before and during hunting seasons in the Bennett
Hills, Id a h o ...................................................................................................... 34
5. Distances, in meters, from random locations and deer locations overall,
pre and post 15 August, and between years, to the nearest water and
livestock in the Bennett Hills, Idaho.......................................................
36
6. Sources of mule deer bed shade, the percent of all beds each shaded,
and sample sizes in the Bennett Hills, Idaho ................................................ 37
7. Mean percent relative density and standard deviation of discerned
fragments from mule deer fecal samples collected in the Bennett
Hills, Idaho ...................................................................................................... 37
8. Frequency of utilization greater than 5% and median utilization class
of selected shrubs and forbs noted at 383 deer locations that did not
have fresh sign of livestock ...........................................................................
38
9. Independent variables and their summary statistics (Average % cover and
standard deviation) chosen for inclusion in the model by both subset and
stepwise analysis of the logistic regression for predicting deer (n=489) and
random (n=284) locations in the Bennett Hills, Id ah o .................................. 39
LIST OF TABLES - continued
Frequency (percent of plots with plant species)/median/mean cover classes
of vegetation recorded at deer locations, by cover type, on the Bennett
Hills study area, 1993 and 1994 .................................................... .............
Frequency (percent of plots with plant species)/median/mean cover classes
of vegetation recorded at random locations, by cover type, on the Bennett
Hills study area, 1993 and 1994 ..................................................................
X.
LIST OF FIGURES
Figure
1. Map of the Bennett Hills Study Area. Heavy lines represent roads.
Lighter lines depict drainages........................................................................
Page
7
2. Average maximum temperature comparisons by month for 1993,1994,
and the 45-year mean, as measured at the Fairfield Ranger Station.............. 17
3. Average precipitation comparisons by month for 1993,1994, and the
45-year mean, as measured at the Fairfield Ranger Station..........................
17
4. Median phenology of Prunus spp. shrubs as measured at deer locations
in 1993 and 1994 in the Bennett Hills, Id ah o ............................ ...................
18
5. Median phenology of Chrvsothamnus viscidifiorus as measured at deer
locations in 1993 and 1994 in the Bennett Hills, Id a h o ................................
18
6. Median phenology of Agastache sp. forbs as measured at deer locations
in 1993 and 1994 in the Bennett Hills, Id ah o ................................................ 19
7. Home ranges of female deer in the Bennett Hills, Idaho in 1993 and 1994:
a) deer 360, b) deer 430, c) deer 510, d) deer 611 ........................................ 21
8. Home ranges of female deer in the Bennett Hills, Idaho in 1993 and 1994:
a) deer 771, b) deer 1140, c) deer 1560, d) deer 1800, e) deer 1950............. 22
9. Overall deer use and availability of cover types in the Bennett Hills,
Id a h o ................................................................................................................ 23
10. Availability, feeding use, and bedding use of cover types by mule deer in
the Bennett Hills, Idaho ................................................................
11. Availability, use in 1993, and use in 1994 of cover types by mule deer in
the Bennett Hills, Idaho .................................................................................. 26
12. Availability and overall use of aspects by mule deer in the Bennett Hills,
Id ah o ................................................................................................................ 26
24
xi
LIST OF FIGURES - continued
Figure
13. Availability and use in 1993 and 1994 of aspects by mule deer in the
Bennett Hills, Idaho.......................................................................................
Page
27
14. Availability and use of aspects within the Mountain Brush cover type
by mule deer in the Bennett Hills, Idaho ...................................................... 28
15. Availability and use of topography by mule deer in the Bennett Hills,
Id ah o .......................................................... ...................................................
29
16. Availability and use in 1993 and 1994 of topography by mule deer in
the Bennett Hills, Idaho ................................................................................ 30
17. Availability and overall use of horizontal configurations by mule deer
in the Bennett Hills, Idaho ............................................................................ 31
18. Availability and use in 1993 and 1994 of horizontal configurations by
mule deer in the Bennett Hills, Idaho............................................................ 32
Xll
ABSTRACT
The factors governing mule deer (Odocoileus hemionus) habitat use patterns in
south-central Idaho are poorly understood. Therefore, 16 radio-collared does were
tracked from the ground in 1993, and 9 of the same does were also tracked in 1994 to
gather data on habitat use and to document the effects of the primary land use in the area,
cattle grazing, on mule deer. The field season of 1993 was cooler and wetter than the
1994 season. Average home range size during the hot and dry field season was
significantly smaller than the average home range size during the cool, wet season
(PO .OI). Deer showed a strong preference for the Mountain Brush cover type for all
activities. Aspen and Riparian cover types for bedding, and Sage/Bitterbrush for feeding.
Cooler northeastern and eastern aspects were preferred by deer and southwestern and flat
aspects were avoided. Overall, deer appeared to prefer steeper than average areas, but
within Mountain Brush they chose flatter sites. Areas near water were preferred and
stock ponds may have made bedding areas that would have been too far from water
otherwise available for use. Deer preferred microsites that did not have fresh sign of
cattle. Habitat use appeared to overlap more between deer and cattle during the dry year
and the latter part of the summer/fall period. A more detailed analysis of the effects of
cattle on mule deer was not possible due to the high density of different landowners in
the area. Hunting season did not affect bed concealment or flushing distance of deer but
may have caused movements out of established home ranges and premature migration to
the winter range. Deer chose shaded areas for bedding and preferred clumps of Salix sp.
shrubs within the Mountain Brush cover type. Microhistological fecal analysis
documented the transition from forbs to shrubs in late summer. A model built with
multiple logistic regression identified overhead cover as the most important factor for
determining sites likely to be used by deer. This study documented the importance of the
Mountain Brush cover type to mule deer in sage steppe environments. It also identified
the potential of hunting pressure to trigger premature migrations.
I
INTRODUCTION
Although mule deer fOdocoileus hemionus l are one of the most studied animals in
North America, much remains to be learned in order to accomplish desired management
goals. Much of the shortage of information can be attributed to the difficulty in
measuring interactions between an animal and its environment. Add to this the broad
spectrum of habitat use strategies utilized by mule deer to cope with conditions as varied
as forested mountains with severe winters to hot, dry deserts in Mexico, and it becomes
evident why many questions remain unanswered. This study was initiated as part of a
long range program (Unsworth 1992) to answer questions about habitat factors which
govern mule deer in the high elevation dry steppe of south-central Idaho and how these
factors are related to livestock grazing, the primary land use of unirrigated lands in the
'
steppe ecosystem.
Studies of mule deer habitat use (Mackie 1970, Uresk and Uresk 1982, Fielder and
Mckay 1984, Austin and Umess 1985, Carson and Peek 1987, Kufeld et al. 1988)
indicate that the habitat features critical to mule deer vary among sites. Factors
influencing habitat use included water, topographical features, horizontal configuration,
aspect, slope, elevation, roads, human activities (mainly hunting), cover types, forage,
and livestock. Many of these factors are inter-related, making it difficult to identify
individual effects.
Rautenstrauch and Krausman (1989) indicated that availability of water was the
primary factor affecting mule deer habitat selection in southwest Arizona. Mackie
2
(1970) found decreased use of areas greater than 3/4 mile from water during summer and
fall in central Montana. He also showed that deer preferred steeper, more rugged terrain
than randomly available. Roads have been found to influence habitat use by mule deer
(Rost and Bailey 1979). Brown and McDonald (1988) reported decreased diurnal
activity and a shift to more secure habitats during hunting seasons in southeast Idaho.
Austin and Umess (1985) discussed the importance of forage availability in an area with
limited summer habitat. Succession may also play a large role in mule deer habitat
selection (Bodurthaetal. 1989).
Ragotzkie and Bailey (1991) and Loft et al. (1991) found that livestock grazing had
significant impacts on mule deer habitat use. In years of average and above average
precipitation, mule deer and cattle diets have little overlap in many regions (Mackie
1970, Currie et al. 1977). This overlap probably increases during years of drought (Short
1977) or when overgrazing limits forage. In spite of this, mule deer have been shown to
prefer ungrazed pastures even in years of adequate forage production (Loft et al. 1991,
Ragotzkie and Bailey 1991).
Moderate grazing by cattle on mule deer summer range in the Sierra Nevadas was
shown to decrease the availability of hiding cover for deer in meadow-riparian and aspen
habitats (Loft et al. 1991), to reduce use of preferred habitats by deer (Loft et al. 1991), to
increase home range size of deer (Loft 1988), and to increase the time deer spent feeding
compared to areas with no grazing (Kie et al. 1991). These factors could adversely affect
mule deer through increased exposure to predators and increased costs of foraging.
3
Although negative effects are abundant in the literature, cattle grazing may be an
effective tool for enhancing forage availability for mule deer. Willms et al. (1979) found
increased spring mule deer use of a pasture grazed the previous fall. On 2 adjacent
ungrazed pastures in central Arizona, Wallace and Krausman (1987) reported higher deer
densities in the pasture that was grazed the previous year. A rest-rotation grazing system
is used on several elk winter ranges managed by the Montana Department of Fish,
Wildlife and Parks to enhance forage (Frisina pers. comm. 1992). Livestock grazing may
also promote diversity in habitat (Mackie 1978).
Despite the potential detrimental effects of cattle grazing on mule deer, cattle
grazing will likely continue to be a dominant aspect of public land use in the West, and
mule deer will likely continue to justify their presence on public land. Mule deer are
Idaho's most important big game animal, providing over 1,000,000 days of recreation in
1991 (Unsworth 1992). In a bioeconomic analysis Loomis et al. (1991) found that the
incremental benefits of deer hunting gained under a 2-years-off I-year-on grazing system
are greater than the lost net economic value of the forage to the rancher. Managers need
to develop methods for management of both deer and cattle on the same lands.
The deer studied were part of a large migratory herd that moved from 20 km to 65
km between their summer range in the Bennett Hills and Sawtooth mountains and their
winter range near King Hill, Idaho. The primary objective of this study was to describe
summer/fall habitat use patterns of female mule deer in the Bennett Hills by quantifying
variables discussed in the literature as possible factors influencing habitat use. Ancillary
4
objectives included documenting the effects of livestock, hunting, and seasonal and
annual variation in temperature and precipitation on female mule deer in the study area.
5
STUDY AREA
The study area was located in the Bennett Hills of south-central Idaho, in Big Game
Management Unit 45 of Idaho Department of Fish and Game (IDFG) Region 4. It
included portions of Elmore, Camas, and Gooding counties. The Bennett Hills are
bounded on the north by Camas Prairie and on the south by the Snake River floodplain.
Most of the study animals summered on the northern front of the Bennett Hills.
Elevations in this geological formation range from 1,500 m in the south portion to
2,100 m on Bennett Mountain.
Located on the southern edge of the Idaho Batholith, much of the soil in the Bennett
Hills is derived from granite. Snake River Basalt is also common. Most of the basalt has
a mantle composed of loess and/or alluvium, but in some areas the mantle is absent.
Andesite and rhyolite are also common components of some soils in the Bennetts (Case
1981, Noe 1991).
Mean annual precipitation at the Fairfield Ranger Station, near the study area, is
40.5 cm. Approximately 70% of the annual precipitation falls from November through
April in the form of snow. Mean annual daily minimum temperature is 3° Celsius and
maximum temperature is 14° C (National Climatic Data Center 1993).
The Bennett Hills are primarily rolling sagebrush rangeland with scattered stands of
aspen and brush. Conifers were very rare on the study area. Riparian zones, consisting
primarily of 3-4 m tall willow ISalix sp.), bordered the perennial streams, Camas Creek,
6
Sheep Creek, King Hill Creek, and Dempsey Creek, on the study area. Meadows were
common in level, low-lying areas.
The primary land use in the Bennett Hills is livestock grazing. Domestic sheep pass
through the area in late spring and cattle are present all summer and fall. Ownership on
the study area was predominantly private (71%), but there were also significant amounts
of land managed by the Bureau of Land Management (9%), and Idaho Department of
State Lands (17%). There was also a small portion of United States Department of
Agriculture Forest Service land (3%).
• Cat
• Deer
iuafain
• Mobnt Bennnett Hills
Figure I. Map of the Bennett Hills Study Area. Heavy lines represent roads. Lighter lines depict drainages.
8
METHODS
Capturing and Radio-collaring Mule Deer
Forty-three mule deer were captured with a net gun from a helicopter and radiocollared during the winter of 1992. Each deer was fitted with a transmitter that had a
time-delay, mortality-sensing device. Age was determined at capture from tooth
replacement and wear (Robinette et al. 1957). Of the 22 radio-collared mule deer that
survived the winter, only 5 (3 does, 2 bucks) remained in the Bennett study area. The
remainder moved further north into the Sawtooth mountains. One of the bucks could not
be tracked from the ground due to an uncooperative landowner. Thirteen additional does
were captured and collared on July 11,1993 to provide an adequate sample. Additional
bucks could not be collared in the summer due to their antler growth. Since I buck is not
an adequate sample from which to infer habitat use, only the 16 does were intensively
tracked in 1993.
Thirty-four more deer were captured and radio-collared during the winter of 1993.
Of the 31 radio-collared deer that survived the winter, only 3,(1 doe, 2 bucks), remained
Z
in the Bennett study area. These could not be tracked from the ground due to an
uncooperative landowner so the 9 does remaining from the collaring efforts in 1993 were
again tracked in 1994.
9
Deer Relocations
Radio-collared mule deer were relocated from a fixed wing aircraft (Maule M5235) bi-monthly during the non-hunting season and weekly during September, October
and November when weather conditions and pilot schedules allowed. The coordinates of
each relocation were recorded by an onboard computer using a geographical positioning
system. Information recorded included cloud cover, ambient temperature, relative wind
speed, habitat, aspect, date and time. When a mortality signal was received, the radiocollared deer was located from the ground and cause* of death determined. Aerial
locations were used only for calculating home ranges.
Error of aerial locations can vary from less than 100 m to I km (Patric et al. 1988)
so intensive ground tracking was employed to attain precise locations for use in micro
level habitat analysis. A 2-element, H antenna and scanner/receiver were used to locate
deer. All tracking was done during daylight hours. I attempted to obtain visual
observations whenever possible. Locations were recorded as I) visual, 2) auditory, or
3) triangulation. Auditory locations were common in dense brush when the deer would
flush at close range but remained unseen due to screening vegetation. Triangulation
locations were estimated from 3 bearings taken from as close as practical and with the
greatest possible separation of angles to reduce the size of the error polygon (Heezen and
Tester 1967, Springer 1979). These were not used for habitat analysis. Each relocation
was plotted on topographic maps and recorded in latitude and longitude to 0.01 minute
using a Garmin Global Positioning Device. Locations were later converted to Universal
10
Transverse Mercator (UTM) coordinates for home range analysis. Activity, group size,
number of fawns, and number of deer seen within I km of the radio-collared animal were
recorded. Activity was classified as feeding, bedding, traveling, escape, or unknown. At
each location where a bed was found, concealment, percent of the bed shaded, what the
bed was shaded by, and the distance I was from the deer when it flushed were recorded.
Concealment was represented by an estimation of the percent of the deer not visible at 50
feet (17 m). A concealment rating scale from I to 10 was constructed by dividing the
percent of the deer not visible by 10.
At each location where a radio-collared deer was seen or sign was found following
an auditory location, the following site characteristics were recorded: elevation, aspect,
slope, topography, horizontal configuration, and distance to the nearest water, stock, and
primary and secondary roads. Elevation was taken from topographic maps
(scale=l :24,000). Slope was measured with a clinometer. Topography was recorded as
ridge top, upper slope, mid slope, lower slope, stream bed/bottom, or bench/flat.
Horizontal configuration was classified as convex, straight, concave, or undulating. The
nearest water was categorized as natural or artificial (i.e. stock ponds). I also noted if the
pastures were stocked and if fresh sign of livestock were found within the plot.
Maintained gravel roads were categorized as primary while tracks or unmaintained
gravel roads were categorized as secondary. Distances to roads and water were measured
from topographic maps (scale=1:24,000).
Cover type was also recorded at each location. Seven major cover types
were identified by vegetation composition and structure. They were: Meadow, Low
11
Sage (<1 m), High Sage (>1 m), Sage/Bitterbrush, Mountain Brush, Riparian, and Aspen.
The Meadow type was found in low-lying areas which remained wet into late
spring. It was composed of a grass/forb community and accounted for relatively little of
the total area.
Big sagebrush (Artemisia tndentata) habitats dominated most of the area.
Bitterbrush (Purshia tndentata), snowberry (Svmnhori carpos sp.), currant (Ribes spp.),
rabbitbrush (Chrvsothamnus nauseosus and C viscidiflorusl and serviceberry
(Amelanchier alnifolial were important shrub components of sagebrush habitats.
Common forbs and grasses included balsamroot (Balsamorhiza spp.), buckwheat
(Eripgonum spp.), lupine (Lupinus spp.), Idaho fescue (Festuca idahoensisl needlegrass
(Stipa spp.), and oniongrass (Melica spp.).
Thick stands of bitter cherry (Prunus emarginatal and chokecherry (R yirginiana)
were scattered throughout the area. Shrub components found in association with the
cherries included Oregon grape (Bejberis repensl ceanothus (Ceanothus sp.), willow
(Salix spp.), stunted quaking aspen (Populus tremuloidesl snowberry, and serviceberry.
There was generally little grass or forb growth in the Mountain Brush habitats.
The Riparian type was found along streams and consisted primarily of very dense
stands of willow shrubs. Mint (Mentha spp.) and sedge (Carex spp.) were commonly
associated with these wet areas. In some areas cattle had reduced or eliminated the
willow cover.
Homogeneous stands of aspen were scattered throughout the area. Stands heavily
utilized by cattle tended to have very little understory, whereas those more remote from
12
cattle had understories of variable density consisting of chokecherry, currant,
serviceberry and snowberry. Forbs and grasses tended to be relatively sparse but
horsemint fAgastache sp.) was commonly encountered.
At deer locations, all vegetation present within a 375 m2 circular plot was noted.
The percent cover for each species was estimated and assigned to a cover class (1=0-1%,
2=>l-5%, 3=>5-25%, 4=>25-50%, 5=>50-75%, 6=>75-95%, 7=>95%). In addition, the
phenologic stages (West and Wein 1971) of shrubs, forbs, and grasses were recorded
(Table I). Utilization of each species was estimated and assigned to a utilization class
(1=0-5%, 2=>5-20%, 3=>20-40%, 4=>40-60%, 5=>60-80%, 6=>80%). Eveiy 2 to 3
days during the 1993 field season, fresh deer droppings were picked up, labeled, and put
in paper bags. These were later analyzed microhistologically by the Composition
Analysis Laboratory in Fort Collins, Colorado.
Table I. Phenological stages and codes used for vegetation at Bennett Mountain.______
Phenology Code
I
2
3
4
5
6
Shrubs
Flower
Fruit set
Fruit swelling
Fruit turning color
Fruit ripe
Fruit dry/dropping
Grasses and Forbs
Vegetative growth
Flower buds
Flower
Fruit set
. Fruit swelling
Plant curing
Random Locations
Habitat availability was determined by going to 299 random points on the study
area and measuring all habitat characteristics recorded at deer locations. This was
accomplished by creating random locations in latitude and longitude to 0.01 minute
13
using a random number table. These locations were then entered into a Garmin Global
Positioning Device which was used to locate the random points with m i n i m a l human
bias. To confirm that an adequate sample size of random plots was done, summary
statistics were computed for 6 variables with subsets of the random points in increments
of 25. The means or proportions of all variables tested stabilized by at least the 200th
random point.
Home Range Analysis
Home range, "that area traversed by the individual in its normal activities of food
gathering, mating and caring for young" (Burt 1943), was calculated and plotted for each
radio-collared deer from June until the deer returned to winter range in 1993 and 1994.
The minimum convex polygon (MCP) (Mohr 1947, Jennrich and Turner 1969) and the
adaptive kernel methods (Worton 1989) were used to compute home ranges. The
CALHOME microcomputer program ( Kie et al. 1994) was used to run these analyses.
Ground and aerial locations were used to calculate home ranges. Home ranges were not
calculated for deer with <20 locations.
Statistical Analyses
Univariate Analyses
Student's t-tests (Freedman et al. 1978) were used to compare group means of all
continuous variables such as slope, elevation, concealment, distance jumped, and
distances to roads, water and stock. Groups that were compared included deer locations
14
versus random locations, deer locations prior to 15 August versus deer locations after 15
August, deer locations prior to hunting season versus deer locations during hunting
season, and 1993 deer locations versus 1994 deer locations. Since the deer tracked in
1994 were also tracked in 1993, a paired t-test was used to compare home range sizes
between years. Overall comparisons included all radio-collared deer, whereas
comparisons between years included only deer tracked both years.
For categorical variables such as cover type, aspect, topography and horizontal
configuration, chi-square, goodness of fit tests (Zar 1984), were used to compare use
versus availability and use in different time periods. Variables which showed significant
variation from expected were subjected to the procedure suggested by Marcum and
V
Loftsgaarden (1980) to determine preference, avoidance, or use in proportion to
availability of individual categories by constructing Bonferroni simultaneous confidence
intervals. This test is not as strong as the goodness of fit procedure because estimates of
availability reduce the power of the test. MSUSTAT (Lund 1993), version 5.2, was used
to run the t-tests, chi-square tests, and to produce summary statistics.
Multivariate Analysis
Multiple logistic regression, using best subset analysis (Hosmer and Lemeshow
1989), was used to create a model that identified a set of habitat variables that
maximized the probability of predicting whether a given site was a deer location or a
random location. The data set was also subjected to the step-up procedure. Only
variables significant at the 0.15 level were included in the step-up model. The 7
15
independent variables used in the analysis were desirable food forb cover, desirable food
shrub cover, lateral hiding cover, overhead hiding cover, slope, and distances to the
nearest road and water (Table 2). These variables were chosen because they, were
significant in univariate analyses. The dependent variable had values of 0 for random
points and I for deer locations. SAS, version 6.10, was used to run this analysis (SAS
Institute 1989).
Table 2. Description of variables used in the multiple logistic regression.
Variable
Description
1. Forb Cover
Combined cover of forbs known to be utilized by mule deer,
including: Castilleia applegatei. Lupinus leucophvllus.
Sphaeralcea grossulariifolia. and Potentilla sp.
2. Shrub Cover
Combined cover of shrubs known to utilized by mule deer,
including: Prunus spp., Ceanothus sp., Rosa sp., and Purshia
tridentata.
3. Lateral Cover
Combined cover of plants providing lateral cover, including:
Ceanothus sp., Svmphoricarpos sp., Ribes spp., Artemisia
tridentata. Purshia tridentata. and Chrvsothamnus nauseosus.
4. Overhead Cover
Combined cover of plants providing overhead cbver, including:
Populus tremuloides. Salix spp., Prunus spp., and Amelanchier
alnifolia.
5. Slope
Slope, measured in degrees.
6. Road Distance
Distance, in meters, to the nearest primary or secondary road.
7. Water Distance
Distance, in meters, to the nearest source of water.
16
RESULTS
In 1993, the 16 radio-collared does were relocated from the ground a total of 346
times throughout summer and fall. Visual observations were made on 77.7% of the
relocations and 11.0% were established by hearing the animal in the brush. The
remaining 11.3% of the ground locations were based upon triangulation and were used
only for home range calculations. The 105 aerial locations logged during the same
period were also used only in home range calculations.
Nine of the same deer were relocated 270 times from the ground throughout'
summer and fall of 1994. These locations were 58.9% visual, 20.7% auditory, and
20.4% triangulation. Thirty-one aerial locations were used in home range calculations.
Climatic Differences Between Years
The field season of 1993 was characterized by unusually cool, wet conditions,
whereas the 1994 season was hot and dry (Figures 2 and 3). Phenology of plants as
measured at deer locations progressed faster in 1994 (Figure 4, 5 and 6). The differences
in years allowed me to compare deer habitat use in a dry year versus a wet year. For all
measured variables that may have been affected by this, between years comparisons have
been included. The 1993 data used in between years comparisons are only from the 9
deer also tracked in 1994.
17
-T
TEMPERATURE (CELSIUS)
35
June
Ju ly
A ug u st
S e p te m b e r
m
M ONTH
Average maximum temperature comparisons by month for 1993, 1994, and the
45-year mean, as measured at the Fairfield Ranger Station.
Legend
I
I
1993
1994
PRECIPITATION (CM)
M ean
June
Ju ly
A ugust
M ONTH
Figure 3. Average precipitation comparisons by month for 1993, 1994, and the
45-year mean, as measured at the Fairfield Ranger Station.
PHENOLOGY
18
T IM E IN T E R V A L S
PHENOLOGY
Median phenology of Prunus spp. shrubs as measured at deer locations in 1993
and 1994 in the Bennett Hills, Idaho (phenology codes are in Table I).
6 / 1 - 6 /1 5
6 /1 6 - 6 /3 0
7 /1 - 7 /1 5
7 /1 6 - 7 /3 1
8 / 1 - 8 /1 5
8 /1 6 - 8 /3 1
9 /1 - 9 /1 5
9 / 1 6 - 9 /3 0 1 0 / 1 - 1 0 / 1 5
T IM E IN T E R V A L S
Figure 5. Median phenology of Chrvsothamnus viscidiflorus as measured at deer
locations in 1993 and 1994 in the Bennett Hills, Idaho (phenology codes are
in Table I).
19
Legend
1993
1994
6 /1 -e /1 5
6 /1 6 - 6 /3 0
7 /1 - 7 /1 5
7 /1 6 - 7 /3 1
8 / 1 - 8 /1 5
8 /1 6 -8 /3 1
9 / 1 - 9 /1 5
9 / 1 6 - 9 /3 0 1 0 / 1 - 1 0 / 1 5
T IM E IN T E R V A L S
Figure 6. Median phenology of Agastache sp. forbs as measured at deer locations in
1993 and 1994 in the Bennett Hills, Idaho (phenology codes are in Table I).
Home Range Analysis
Summer/fall home ranges were calculated for 9 deer tracked in 1993 and 1994 and
for 6 additional deer tracked only in 1993 (Table 3). One deer was excluded from
analysis because of an insufficient number of locations (n<20). A total of 439 locations
were used in the calculation of 1993 home ranges, while 301 were used in 1994. Home
ranges included locations from the time the deer arrived on the summer range in early
June until they began moving back to the winter range in late October.
The minimum convex polygon method yielded smaller home range estimates than
the 95% adaptive kernel in 87.5% of the data sets analyzed. Home ranges calculated
20
with the MCP averaged 3.59 km2 (SD= 1.73 km2) whereas those calculated with the 95%
adaptive kernel averaged 4.22 km2 (SD=1.97 km2).
Table 3. Numbers of locations and home range sizes for all deer tracked in the Bennett
_______ Hills, 1993 and 1994.
,
Ground
Locations
Aerial
Locations
Total
Locations
MCP
(km2)
95% Adapt.
Kernel (km2)
Deer
93
94
93
94
93
94
93
94
100
19
6
25
4.66
310
18
8
26
5.01
730
24
12
36
4.84
1391
19
6
25
4.94
1820
23
6
29
7.63
1940
19
9
28
179
360
24
29
8
3
32
32
185
1.72
430
25
30
8
4
33
34
5.12
2.54
510
23
34
8
5
31
39
4.39
2.26
611
23
29
7
3
30
32
4.44
0.81
771
22
32
2
4
24
36
5.75
172
1140
23
7
28
3
30
31
5.65
1.28
1560
25
6
29
3
31
32
2.73
2.28
1800
23
31
5
3
28
34
3.16
1.88
1950
24
7
28
3
31
31
0.75
2.92
Mean* 23.6
30.0
6.4
3.4
30.0
33.4
4.00
2.14
*1993 means are derived from only the 9 deer tracked in 1994 also.
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
93
4.46
7.54
6.78
192
8.00
4.74
4.03
5.57
5.21
165
4.41
7.26
122
4.11
3.19
4.52
94
-
_
-
-
-
2.01
109
3.72
0.90
132
1.69
3.08
2.53
192
2.58
Home ranges of the 9 deer tracked during both field seasons were significantly
larger in 1993 (PO.Ol). Estimates using the 95% adaptive kernel method averaged 4.52
km2 (SD =1.31 km2) in 1993 and 2.58 km2(SD = 1.41 km2) in 1994. The MCP home
range estimates averaged 4.00 km2 (SD = 0.99 km2) in 1993 and 2.14 km2 (SD = 1.50
km2) in 1994. In all cases the number of locations used in 1994 calculations equalled or
21
exceeded the number used in 1993. Number of locations per deer averaged 30.0 for 1993
calculations and 33.4 for 1994 calculations. The spatial relationship of individual deer's
home ranges between years can be seen in Figures 7 and 8.
(b)
(c)
Figure 7. Home ranges of female deer in the Bennett Hills, Idaho in 1993 and 1994:
a) deer 360, b) deer 430, c) deer 5 10, d) deer 611.
22
(c)
(d)
(e)
Figure 8. Home ranges of female deer in the Bennett Hills, Idaho in 1993 and 1994 :
a) deer 771, b) deer 1140, c) deer 1560, d) deer 1800, e) deer 1950.
23
Cover Type Use
Overall Cover Type I Jse
Cover type was recorded at 535 deer locations during the summer/fall field seasons
of 1993 and 1994 (Figure 9). With all activities and both years combined, deer did not
use cover types in proportion to availability (X2=I 879, d.f =6, PO .OI). The MarcumLoftsgaarden analysis indicated that deer selected for Mountain Brush and Aspen,
selected against Grass, Low Sage, and High Sage, and used Sage/Bitterbrush and
Riparian cover types in proportion to availability (P<0.05).
GR
LS
HS
BB
MB
Rl
AS
COVER TYPES
Figure 9. Overall deer use and availability of cover types (GR=Grass, LS=Low Sage,
HS=High Sage, BB=Bitterbrush, MB=Mountain Brush, RI=Riparian,
AS=Aspen) in the Bennett Hills, Idaho.
24
Activity and Cover Use
Cover type was recorded at 396 deer bedding sites, 82 deer feeding sites, and 57
unknown activity sites (Figure 10). The locations classified as unknown were mainly
instances where the deer was not seen or heard until it was fleeing and subsequent
inspection of sign in the area did not indicate the type of activity.
Deer did not use cover types for bedding in proportion to availability (X2=ITOS,
d.f.=6, P O .01). The Marcum-Loftsgaarden analysis indicated that deer selected for
Mountain Brush, Riparian, and Aspen and selected against Grass, Low Sage, High Sage,
and Sage/Bitterbrush cover types for bedding (PO .05).
60
L eg en d
Availability
50-
Feeding
Bedding
COVER TYPES
Figure 10. Availability, feeding use, and bedding use of cover types (GR=Grass,
LS=Low Sage, HS=High Sage, BB=Bitterbrush, MB=Mountain Brush,
RI=Riparian, AS=Aspen) by mule deer in the Bennett Hills, Idaho.
25
Deer also did not use cover types for feeding in proportion to availability
(X2=123.2, d.f.=6, P<0.01). The Marcum-Loftsgaarden analysis indicated that deer
selected for Sage/Bitterbrush and Mountain Brush, selected against Grass and High Sage,
and used Low Sage, Riparian, and Aspen cover types in proportion to availability for
feeding (P<0.05).
Habitat use differed significantly between feeding and bedding sites (X2=Sd-O?,
d.f =6, P O .01). The Marcum-Loftsgaarden analysis indicated that deer used
Sage/Bitterbrush more for feeding, Mountain Brush, Riparian and Aspen more for
bedding, and the other cover types equally (PO .05).
Variability Between Years
For deer followed in both years, cover type was recorded at 193 locations in 1993
and 224 locations in 1994 (Figure 11). Cover type selection differed significantly
between years (X2=14.30, d.f.=6, PO.03). The Marcum-Loftsgaarden analysis indicated
that deer used High Sage more in 1993 than in 1994 and used all other cover types
equally in both years (P O .05).
Aspect Use 1
Overall Aspect Use
Aspect was recorded at 522 deer locations during the summer/fall field seasons of
1993 and 1994 (Figure 12). With all activities and both years combined, deer did not use
aspects in proportion to availability (X^=416.2, d..f.=8, P O .01). The MarcumLoftsgaarden analysis indicated that deer selected for northeastern and eastern, selected
26
Legend
III
LU
O
A vailab ility
1993
1994
30
SL
HS
BB
MB
C O V ER TYPE
Figure 11. Availability, use in 1993, and use in 1994 of cover types (GR=Grass, LS=Low
Sage, HS=High Sage, BB=Bitterbrush, MB=Mountain Brush, RI=Riparian,
AS=Aspen) by mule deer in the Bennett Hills, Idaho.
Legend
A vailability
U se
NE
NW
FLAT
A SPECTS
Figure 12. Availability and overall use of aspects by mule deer in the Bennett Hills,
Idaho.
27
against southwestern and flat aspects, and used all other aspects in proportion to
availability (P O .05).
Variability Between Years
Aspect was recorded at 193 deer locations in 1993 and 210 deer locations in 1994
(Figure 13). Aspect selection differed significantly between years ( X 2= \ 6 . 2 6 , d.f.=8,
P O .01). The Marcum-Loftsgaarden analysis indicated that deer used southeastern
aspects more in 1993, northern aspects more in 1994, and all other aspects equally in
both years (P O .05).
35
i
---------------------------------------------- ]--------------------------L eg en d
N
NE
E
SE
S
SW
W
NW
FLAT
ASPECTS
Figure 13. Availability and use in 1993 and 1994 of aspects by mule deer in the
Bennett Hills, Idaho.
28
Relation to Mountain Brush
An analysis of aspect use within Mountain Brush was included because it was the
most used cover type on the study area and cover types were at least partially dictated by
aspect. Aspect was recorded at 249 deer locations in the Mountain Brush cover type
(Figure 14). Aspect selection within Mountain Brush differed significantly from aspect
availability within Mountain Brush (X2=41.69, d.f.=8, P O .01). The MarcumLoftsgaarden analysis indicated that deer selected for northeastern aspects, selected
against southern aspects and used all other aspects within Mountain Brush in proportion
to availability (PO .05).
ASPECTS
Figure 14. Availability and use of aspects within the Mountain Brush cover type by
mule deer in the Bennett Hills, Idaho.
29
Topography Use
Overall Topopraphv Use
Topography was recorded at 521 deer locations during the summer/fall field
seasons of 1993 and 1994 (Figure 15). With all activities and both years combined, deer
did not use topography in proportion to availability (X^=271.8, d.f =5, P<0.01). The
Marcum-Loftsgaarden analysis indicated that deer selected for mid-slope, selected
against lower slopes and benches or flats, and used ridge tops, upper slopes and stream
beds in proportion to availability (P<0.05).
Legend
TO POG RA PH Y
Figure 15. Availability and use of topography (RT=ridge top, US=upper slope, MS=midslope, LS=Iower slope, SB=Stream bed, BF=bench or flat) by mule deer in the
Bennett Hills, Idaho.
30
Variability Between Years
Topography was recorded at 193 deer locations in 1993 and 209 locations in 1994
(Figure 16). Topography use differed significantly between years (X2=IOb^, d.f=5,
P O .01). The Marcum-Loftsgaarden analysis indicated that deer used upper slopes more
in 1993, mid-slopes more in 1994, and all other areas equally in both years (PO .05).
MS
LS
TOPOGRAPHY
Figure 16. Availability and use in 1993 and 1994 of topography (RT=Tidge top,
US=Upper slope, MS=mid-slope, LS=Iower slope, SB=Stream bed, BF=bench
or flat) by mule deer in the Bennett Hills, Idaho.
Horizontal Configuration Use
Overall Horizontal Configuration
Horizontal configuration was recorded at 517 deer locations in the summer/fall
field seasons of 1993 and 1994 (Figure 17). With all activities and both years combined.
31
deer did not use horizontal configurations in proportion to availability (X2= I I l A d.f.=3,
P<0.01). The Marcum-Loftsgaarden analysis indicated that deer selected for straight,
selected against convex and undulating, and used concave horizontal configurations in
proportion to availability (PO .05).
Legend
A vailability
U se
10
C onvex
S tra ig h t
C oncave
IU n d u latin g
H O R IZ O N T A L C O N F IG U R A T IO N
Figure 17. Availability and overall use of horizontal configurations by mule deer in the
Bennett Hills, Idaho.
Variability Between Years
Horizontal configuration was recorded at 193 deer locations in 1993 and 209
locations in 1994 (Figure 18). Horizontal configuration use differed significantly
between years (X2= IIl .4, d.f.=3, P<0.01). The Marcum-Loftsgaarden analysis indicated
that deer used convex more in 1993, straight and undulating more in 1994, and concave
horizontal configurations equally in both years (P O .05).
32
-inn
C onvex
S tra ig h t
C oncave
U n d u latin g
H O R IZ O N T A L C O N F IG U R A T IO N
Figure 18. Availability and use in 1993 and 1994 of horizontal configurations by mule
deer in the Bennett Hills, Idaho.
Elevation and Slope Use
Elevation, as measured at 521 deer locations, averaged 2,039 m (SD=41 m). This
was significantly different than the average of 2,023 m (SD=45 m) measured at random
locations (PO.Ol).
Slope, as measured at 521 deer locations, averaged 12.3° (SD=5.7°). This was
significantly steeper than the average of 8.9° (SD=7.3°) measured at random locations.
Slope at bedding sites averaged 12.2° (SD=5.5°) and did not differ significantly from the
average of 12.6° (SD=5.8°) measured at feeding sites (P=0.55). Slope at deer locations
within Mountain Brush averaged 14.0° (80=4.3°) which was significantly steeper than
deer locations at all other cover types combined which averaged 10.9° (SD=6.3°)
33
(PO.Ol). However, it was significantly less than the average of 16.1° (SD=7.5°)
measured at random locations within Mountain Brush (P=0.01).
Distance to Roads
Overall Distance to Roads
Distance to a primary road, as measured at 494 deer locations, averaged 1,297 m
(SD=841 m). This was significantly less than the average distance of 1,544 m
(SD=999 m) measured at random locations (PO.Ol). Distance to a secondary road
averaged 731 m (SD=517 m) at 516 deer locations which did not significantly differ from
the average of 721 m (SD=576 m) measured at random locations (P O .79).
Before and During Hunting Season
Distance to a primary road was measured at 463 deer locations prior to the opening
of hunting seasons and 31 locations during hunting seasons. Distance to a primary road
averaged 1,278 m (SD-826 m) prior to hunting season and 1,581 m (SD=T,013 m)
during hunting season. This difference approached significance (P= 0514). Distance to
a secondary road averaged 727 m (SD=518 m) prior to hunting season and 790 m
(SD=500 m) during hunting season. This was not a significant difference (P=0.85).
34
Table 4. Distances, in meters, to primary and secondary roads from random and deer
locations, overall, before and during hunting seasons in the Bennett Hills,
Idaho.
Random
Overall
Pre-Hunting
During
Hunting
Primary Rd.
1544
1297
1278
1581
Secondary Rd.
721
731
727
790
Distance to Water
Overall Distance to Water
Distance to water, as measured at 518 deer locations, averaged 364 m (SD=321 m).
This was significantly less than the average distance of 426 m (SD=311m) measured a t .
random locations (PO.Ol). The nearest water source was artificial at 39.7% of deer
locations and 27.1% of random locations. Type of water was not distributed
proportionately between deer and random locations (X2=?13.28, d.f.=T, PO.Ol).
Variability Between Years
The inter-year distances to water did not differ significantly (PO . 56). Mean
distance to water in 1993, based on 191 locations, was 399 m (SD=424 m) and in 1994,
based on 208 locations, was 380 m (SD=202 m). Type of water nearest deer locations
also did not differ between years (X20 . 355, d.f.=T, PO . 12).
35
Variability Between Seasons
Distance to water, as measured at 242 deer locations prior to 15 August, averaged
324 m (SD=224 m). This was significantly less than the average of 398 m (SD=384 m)
measured at 276 locations after 15 August (PO .01).
Distance to Livestock
Overall Distance to Livestock
Distance to livestock, as measured at 483 deer locations, averaged 500 m
(SD=330 m). This did not differ significantly from the average of 446 m (SD=550 m)
measured at random locations (P=OJ I). Fresh sign of livestock was noted at 26.8% of
deer locations which significantly differed from the 43.9% of plots with fresh sign at
random locations (X2=23.89, d.f.=l, PO.Ol). '
Variability Between Years
Distance to livestock was measured at 303 deer locations in 1993 and 180 locations
in 1994. The 1993 average, 570 m (SD=342 m), was significantly larger than the average
of 382 m (SD=272 m) measured in 1994 (P<0.01). Fresh sign of livestock was noted at
24.2% of deer locations in 1993 and 30.3% in 1994. This difference was not significant
(X2=2.272, d .f=1, P=0.13).
Variability Between Seasons
Distance to livestock, as measured at 212 deer locations prior to 15 August,
averaged 561 m (SD=405 m). This was significantly larger than the average of 452 m
36
(SD=247 m) measured at deer locations after 15 August (PO. 01). Fresh sign of
livestock was noted at 20.5% of deer locations prior to 15 August and 32.1% after 15
August. This difference was significant (X^=8.428, d.f.=l, P O .01).
Table 5. Distances, in meters, from random locations and deer locations overall, pre
and post 15 August, and between years, to the nearest water and livestock in the
_______ Bennett Hills, Idaho.______
Random
Overall
Pre 8/15
Post 8/15
1993
1994
Water
426
364
324
398
399
380
Livestock
446
500
561
452
570
382
Concealment and Distance When Flushed
The concealment rating of deer averaged 9.2 (SD=1.2) prior to hunting season and
8.9 (SD= 1.7) during hunting season. This was not a significant difference (P=0.30). My
distance from deer when they flushed also did not differ significantly before and during
hunting season (P=0.75). Prior to hunting season the deer flushing distance averaged
50 m (SD=31 m), and the distance during the season averaged 48 m (SD=30 m).
Deer Bed Shading
At 325 deer bedding locations, the percent of the bed that was shaded was
estimated. Shading averaged 81% (SD= 19%). The primary source of the shade was also
recorded. The 5 most common sources of shade are listed in Table 6. Although all
sources listed in the table are living plants, other less common sources included boulders,
dead horizontal tree trunks, and topographic irregularities.
37
Table 6. Sources of mule deer bed shade, the percent of all beds each shaded, and
sample sizes in the Bennett Hills, Idaho.
Shade Source
% of Beds
n
Prunus spp.
40
135
Ponulus tremuloides
23
78
Salix spp.
20
69
Artemesia tridentata
6
21
Amelanchier alnifolia
4
16
Food Habits
Microhistological fecal analysis was conducted on samples collected in August
through October of 1993 (Table 7). Most of the plants identified in this procedure were
also identified in the field at deer locations (Table 8). When calculating median
utilization classes, I omitted category I (0-5% utilization) because all medians would
have equalled I otherwise.
Table 7. Mean percent relative density and standard deviation of discerned fragments
_______ from mule deer fecal samples collected in the Bennett Hills, Idaho._______
August
Sentember
October
Plants
Mean
SD
Mean
SD
Mean
SD
All Grasses
O
O
0
0
0
0
Seeds
6
7.36
1.26
1.74
‘ 6.07
5.26
Legume Pod
2.46
4.04
4.07
4.36
9.39
5.6
Lupinus spp.
64.89
10.84
61.01
12.59
5.26
2.38
Potentilla spp.
2.89
0.33
1.89
1.73
2.36
2.17
Solonum spp.
0.65
1.46
0
0
0
0
Artemesia tridentata
O
0
0
0 .
0.7
1.57
Berberis repens
O
0
2.51
2.56
1.43
1.96
Ceanothus sp.
2.9
2.96
11.05
2.68
22.59
7.07
Purshia tridentata
18.56
4.52
18.21
8.7
52.2
9.25.
Svmphoricarpos sp.
1.65
1.53
0
0
0
0
38
Table 8. Frequency of utilization greater than 5% and median utilization class (1=0-5%,
2=>5-20%, 3=>20^10%, 4=>40-60%, 5=>60-80%, 6=>80%) of selected shrubs
_______ and forbs noted at 383 deer locations that did not have fresh sign of livestock.
Shrubs
%
Ma Forbs
%
Ma
Acer sp.
75.0 3
Balsamorhiza sp
14.8
2
Amelanchier alnifolia
4.6 2
Castilleia appleuatei
90.9
3
Ceanothus sp.
74.7 2
Crepis spp.
27.7
2
Chrvsothamnus nauseosus
10.0 2
Eriogonum spp.
12.7
2
Ponulus Iremuloidesb
21.0 2
Geranium spp.
8.5
2
Prunus spp.
73.0 2
Helianthella sp
9.0
2
Purshia tridentata
36.7 2
Luninus leucophvllus
100.0 3
Ribes spp.
11.2 2
Penstomen spp.
28.3
2
Rosa sp.
36.7 2
Snhaeralcea erossulariifolia 80.0
4
Salix spp.
58.8 2
Svmnhoricarpos sp
11.0 2
“Medians were calculated by omitting category I (0-5% utilization).
lDnly includes plants < 5m in height.
Multivariate Habitat Analysis
Subset analysis of the multiple logistic regression produced all possible models
\
with combinations of the 7 independent variables. The models were ranked from "best"
to "worst" based on the Cp value (predictive squared error). The best results for
predicting a deer location were derived from a 3 variable model including overhead
cover, shrub cover, and lateral cover (Hosmer Lemeshow statistic =4.159, d.f.=5,
P=0.53):
g - l- 5 2 2 + (0 .0 5 6 * Overhead Cover) + (0.001* Shrub Cover) - (0.013 * Lateral Cover)
Pr(deer) = I + e "1-522 + ^0 056 * Overhead Cover) + (0.001 * Shrub C o v e r ) (0.013 * Lateral Cover)
;
39
Overhead Cover and Shrub Cover were positively related to the probability of a point
being a deer location and lateral cover was negatively related. The overall prediction
success of this model was 39.9%. The best single variable model used only overhead
cover (Hosmer Lemeshow statistic =31.96, d.f.=5, P<0.01) and had a prediction success
of 37.4%. Stepwise analysis of the same data set yielded the same 3 variable model as
the best subset analysis.
Table 9. Independent variables and their summary statistics (Average % cover and
standard deviation) chosen for inclusion in the model by both subset and
stepwise analysis of the logistic regression for predicting deer (n=489) and
_______ random (n=284) locations in the Bennett Hills, Idaho.________________
Deer
Random
Variable
Av SD Av SD
R2
Cna
F
P
Overhead Cover 50
34
9 21
31.96
0.37
461.41
<0.001
Lateral Cover
29
25
40
25
6.67
0.40
27.17
<0.001
Shrub Cover
32
34
7
17
4.16
0.40
4.51
0.034
Cp=predictive squared error.
40
DISCUSSION
Home Range
The overall average MCP estimate of summer/fall home range size for female deer
in the Bennett Hills of 3.6 km2 was similar to that of norimigratory mule deer in
southwestern Texas (3.8 km2; Dickinson and Gamer 1979) and eastern Montana (6.3
km2; Wood et al. 1989), but considerably smaller than the summer/fall means reported in
southcentral Washington (26.9 km2; Eberhardt et al. 1984), westcentral Arizona (32.3
km2; Hayes and Krausman 1993), and southwestern Arizona (121 km2; Rautenstrauch
and Krausman 1989). Ifhome range size is inversely related to habitat quality as
suggested by Bailey (1984), then the Bennett Hills should be excellent habitat for female
mule deer.
Average 95% kernel home range size during the hot and dry field season of 1994
was only 57.1% of the average size during the cooler field season of 1993 (53.5% as
estimated by the MCP) for deer tracked in both years. This may have been the result of
deer concentrating use in more mesic microsites instead of dispersing use over a larger
area. Another factor may have been decreased diurnal activity during the hotter year.
Hayes and Krausman (1993) reported a decrease in diurnal activity by female mule deer
during hot and dry periods that was partially compensated for by increased nocturnal
activity; however, 88% of the area used noctumally fell within areas used during the
daytime. I think it was probably a combination of these 2 factors with the former being
41
the primary explanation. If deer decreased diurnal activity, their bedding areas should
still have been dispersed over similar sized areas in both years.
Cover Type Use
The Mountain Brush cover type was the most frequently used type on the study area
for overall use (48.8%), feeding use (35.4%), and bedding use (52.8%), despite being the
third most common cover type at 11.7% of surface area availability. This preference was
probably due to a variety of factors, including the suitability of Mountain Brush for food,
security and thermal cover. The Mountain Brush cover type was composed primarily of
dense Prunus spp. stands in which evidence of deer browsing was usually conspicuous.
Other important food shrubs such as Oregon grape, ceanothus, and willow were common
in these stands. Although Ibrb growth was generally sparse within the Mountain Brush
cover type, the ecotonal fringe area between the Mountain Brush and adjacent sage cover
types were often characterized by a relatively high density offorbs.
Deer appeared to feel secure in the Mountain Brush cover type. The flushing
distance for bedded deer averaged 45.5 m (SD=26.8 m) in Mountain Brush compared
with 67.2 m (SD=40.7 m) in sage habitats. When locating bedded deer in Mountain
Brush, I often passed by deer at <20 m without being aware of their presence. Deer often
jumped from beds only when I passed by them a second time. Deer bedded in more open
cover types rarely behaved in this manner.
Some deer made repeated use of specific beds in Mountain Brush stands
despite being displaced by me every 3 to 5 days. Repeated use of the same beds over the
42
years created oval depressions worn into the ground. Often a series of traditional beds
would be found around an especially thick clump of brush where deer presumably moved
from bed to bed throughout the day to remain shaded. When locations were made at
these sites, the deer were almost always in the traditional bed providing maximum shade.
Clumps of willow surrounded by Prunus spp. seemed to be especially attractive to deer
seeking a bedding site.
Overall, the Mountain Brush cover type provided a cool, secure place for daytime
activities. In the hotter, dryer summer of 1994, decreased use of the High Sage cover
type compared with the cooler summer of 1993 was noted. Because all locations used
for analyses in this study were obtained during daylight hours, I may have underestimated
use of habitats used at night and overestimated use of habitats used primarily diumally,
such as Mountain Brush. How much bias using only diurnal locations introduced would
be difficult to determine. Hayes and Krausman (1993) reported that general patterns of
diel habitat use were accurately represented by daylight observations. Kufeld et al.
(1988) observed different habitat selection patterns during different periods of the day,
apparently due to different activity patterns between night and day, e.g. increased feeding
at night.
i
The Aspen cover type appeared to be the second most preferred type. Aspen cover
types accounted for 27.0% of deer bedding sites. Mountain Brush and Aspen combined
accounted for 79.8% of bedding locations despite making up only 14.4% of availabile
vegetation. The Aspen cover type was probably preferred for the same reasons discussed
43
for Mountain Brush but to a lesser degree since ground level hiding cover was generally
sparser and availability of preferred food shrubs was lower.
Deer used the Aspen cover type in proportion to availability for feeding. Most
aspen stands in the Bennett Hills had a park-like structure with dense stands of grass
comprising the bulk of the understory. The scarcity of forbs and shrubs probably limited
deer feeding activity. As with Mountain Brush, the areas immediately adjacent to aspen
stands were often characterized by increased shrub and forb production.
Overall, the SageZBitterbrush and Riparian cover types were used in proportion to
availability. However, deer did select SageZBitterbrush for feeding and Riparian for
bedding. I suspect that actual use of SageZBitterbrush for feeding was higher than
daytime sampling indicated. Deer sign was abundant in this type and frequency and
intensity of utilization on bitterbrush was high. Microhistological fecal analysis
indicated high relative density of bitterbrush, especially later in the summer and fall
(52.20% (SD=9.25%) average in October). For these reasons, I suspect that deer utilized
the SageZBitterbrush cover type heavily at night.
The use of the Riparian cover type for bedding was highest during the hottest days
of August. Although no measurements of temperature were taken, I suspect that the
temperature under the tree canopy in the Riparian cover type was the coolest available in
the study area. Extremely dense willow brush shaded damp, lush grass kept cool by
evaporative cooling. However, increased use of this habitat was not noted during the
hotter summer of 1994, so the importance of this habitat during periods of high
44
temperature is not clear. Availability of preferred deer foods in this habitat was generally
low.
The Grass, Low Sage, and High Sage cover types were not preferred by mule deer
under any conditions. The Grass and Low Sage habitats had veiy little to offer to deer in
terms of food or cover. Few preferred deer forbs grew in the Grass cover type, and the
few that did dried out early in the summer due to lack of shade. Most Low Sage was
associated with rock and bare soil flats where all vegetation was sparse, and this
vegetation dried very early in the summer.
High Sage was by far the most abundant cover type on the area, comprising 46.8%
of available vegetation. Deer use of this type was never high, but deer did use it more
during the cooler summer of 1993. Production of forbs and shrubs utilized by deer was
moderately high, and lateral cover for bedding concealment was also good. Sign found
in this cover type indicated that deer used it for feeding at night. The low frequency of
daytime use is probably attributable to the presence of Mountain Brush and Aspen cover
types in the Bennett Hills. In the absence of these highly preferred cover types, I predict
the use of High Sage would greatly increase and/or deer density would be much lower.
Aspect. Topography, and Horizontal Configuration
Overall, mule deer selected for northeastern and eastern aspects and selected
against southwestern and flat aspects. During the hot, dry season of 1994, deer shifted
some use away from southeastern aspects to northern aspects. Deer were probably using
north to east aspects because they were copier. Availability of north, northeast and east
45
aspects combined was 29.7% and deer use of these aspects was 67 J % overall. The
selection of aspect was confounded by the uneven distribution of the Mountain Brush
cover type over aspects. Most Mountain Brush cover types (51.4%) were on north to east
aspects. However, if deer were selecting areas of use based only on the presence or
absence of Mountain Brush then the proportion of aspects used by deer should equal the
proportion of aspects occupied by Mountain Brush. However, this was not the case.
Within Mountain Brush, deer selected north to east slopes disproportionately (81.9%).
They used northeastern aspects significantly more than their frequency of occurrence
within Mountain Brush. Although not statistically significant, it appeared that deer may
have been avoiding southern aspects within Mountain Brush (availability= 17.1%,
use= 1.6%).
Deer used the mid-slope of hillsides more than any other topographic position.
They selected against lower slopes and benches or fiats. Ridge tops, upper slopes and
stream beds were used in proportion to availability. The majority of Mountain Brush
found on the study area was located at mid-slope, which probably explains this selection.
Lower slopes were generally dominated by High Sage and benches or fiats were
primarily covered by Grass and Low Sage habitats, all non-preferred cover types. Deer
used upper slopes proportionately more during the cool 1993 field season and mid-slopes
more during the hot 1994 field season. One possible explanation for this is that mid­
slopes were generally nearer water than upper slopes. A more likely explanation was
that mesic mid-slopes remained cooler and supported better forage than the drier upper
slopes during the dry year.
Deer selected against convex horizontal configurations because convex sites tended
to be dry and unproductive. Explaining the avoidance of undulating sites and the use in
proportion to availability of concave sites is more difficult. Logically, deer should have
preferred these horizontal configurations due to their more mesic nature. I may have
missed nocturnal use of these sites since all locations were diurnal, and/or they may have
been avoided for bedding because they did not provide a clear view of approaching
danger. Convex sites were used proportionately more in the wet year, and straight and
undulating sites were used proportionately more in the diy year. The increased use of
concave sites I expected during the dry year was not observed. This may have been an
artifact of my sampling schedule or due to the reluctance of deer to bed in sites offering
poor visibility.
Elevation and Slope
Although a significant difference between the average elevation of deer and
random locations was detected, the absolute difference (16 m) does not have any
practical implications. The reason for the observed difference was the higher proportion
of random locations on the Camas Creek flats. The deer tended to spend their days on
the slope away from the flats. This tendency was also reflected in the relatively low use
of bench or flat areas (4.2% of deer locations) compared to their availability (22.4% of
random locations).
Deer selected significantly steeper slopes on average than those at random
locations. This, too, was at least partially due to the higher proportion of random
47
locations situated on the flats surrounding Camas Creek. Deer seldom used the flats due
to a paucity of preferred cover types. Mountain Brush was associated with steeper slopes
than most other cover types. Deer apparently selected for Mountain Brush, and steeper
slopes were incidentally associated with that cover type, because slope was steeper at
deer locations within Mountain Brush than in other cover types. Within Mountain Brush,
deer appeared to prefer more level sites than random sampling indicated were available.
Roads. Water and Livestock
No significant relationship was detected between deer and secondary roads.
Neither overall average deer distance compared with random distance nor deer distance
prior to hunting season compared with deer distance during hunting season varied
significantly. Secondary roads were distributed relatively evenly over the study area and
hunter access to these was extremely limited due to private land ownership. Distance to
primary roads was significantly less for deer locations than for random locations. This
was probably an artifact of the location of deer I sampled to the road system in the study
area rather than an actual preference for proximity to primary roads. The primary road
' that was nearest to most points on the study area ran east and west along the northern
border of the study area. Most home ranges of radio-collared deer were near the northern
edge of a remote area extending approximately 24 km south of this road.
The arrangement of primary roads in the study area was also at least a partial
explanation for the greater distance from deer to primary roads measured during hunting
season as compared to before hunting season. At about the beginning of hunting season
48
(October 5) both years, the deer began a gradual movement to the south, towards the
winter range and away from the primary road on the north edge of the study area. This
movement could be interpreted in 2 ways: I) deer were moving away from the primary
road in response to hunting pressure, or 2) deer were simply beginning their seasonal
migration to winter range. I tend to support the latter interpretation because several
radio-collared deer initiated southern movements I to 9 days prior to the opening of
hunting season. The hunting season may have served as a trigger or impetus to the
migration.
Deer locations tended to be closer to water than would be expected, based upon
random locations. Water was not a particularly scarce resource in the Bennett Hills.
Camas Creek, a perennial stream, was located near the northern edge of the study area
and most deer had a portion of the creek in their home ranges. Sheep Creek fed Camas
from the south and ran yearlong in 1993 but dried up in mid-August of 1994. Stock
ponds were distributed throughout the area, and many of them became dry in August of
both years.
A disproportionate number of deer locations were nearer stock ponds than natural
sources of water. This was probably due to the tendency of the flat and gently sloping
areas of the Camas Creek floodplain to support mainly grass and sage communities,
creating a corridor 200-800 m wide of unfavorable deer habitat. Stock ponds were more
commonly located near Mountain Brush and Aspen cover types. Actual utilization of
artificial water sources versus natural water sources cannot be inferred from these data,
but it would be logical to assume that use of stock ponds may increase the fitness of deer
by reducing exposure and travel in low security cover types. The reduction in availability
of water was reflected in the significantly greater average distance after 15 August as
compared to before 15 August. This would imply that as water decreased in availability,
deer did not alter areas of use to maintain a similar proximity to water. I would not
expect this to hold true in areas with extremely scarce water sources.
Overall, average distance to livestock did not vary significantly between deer
locations and random locations; however, deer did appear to prefer sites that did not have
fresh sign of livestock. There was evidence to suggest that habitat use by deer and cattle
overlapped more during dry years and dry seasons than during wetter conditions as
suggested by Short (1977). Both the inter-year and the pre and post 15 August
comparisons showed a significantly shorter average distance to livestock during the dry
year or season. Fresh sign of livestock also significantly increased in frequency at deer
locations after 15 August.
The effects of cattle grazing on mule deer in the Bennett Hills could not be
quantitatively addressed due to the chaotic distribution of land ownership, cattle, and
deer on the study area. The 9 deer studied both years used areas owned by 13 different
private landowners in addition to BLM and Forest Service lands. The home range of
each deer included portions of land owned by >4 different landowners, with each
landowner using different stocking rates and turnout dates. This scenario resulted in no
opportunity to compare deer habitat use in stocked versus non-stocked pastures, or deer
habitat use under varying stocking rates.
50
Concealment and Flushing Distanne
Hunting seasons did not seem to affect either the concealment rating of bedded deer
or the distance I was from bedded deer when they flushed. Evidently deer utilized the
same combination of hiding and escape strategies throughout the year. However,
movements out of the home range area increased during hunting season. Whether this
was in response to hunting season or merely coincidental to the fall migration was
difficult to determine. I detected no evidence of the radio-collared deer becoming
habituated to my approach.
Deer Bed Shading
Sargent et al. (1994) discussed the possible importance of behavioral
thermoregulation by mule deer. They reported that deer on their study area in southcentral Washington sought out shaded bedding sites that still permitted radiative heat
loss. Mule deer in the Bennett Hills behaved similarly. Sources of shade reported for
this study generally reflect the dominant plant species associated with various cover
types; i.e. the most common source of shade, Prunus spp., reflects the most frequently
used cover type for bedding. Mountain Brush, and the second most common source of
shade, Populus tremuloides. reflects the second most used cover type for bedding, which
was Aspen.
Willows did not follow this pattern. They provided primary shade at 20% of the
deer beds while the Riparian cover type in which they dominated accounted for only
J
51
6.3% of deer bedding locations. The explanation lies in the tendency of deer to seek out
isolated patches of willows within otherwise homogeneous stands of Prunus spp. in
Mountain Brush habitat. The denser, taller foliage of the willow may have provided
superior shade and security.
Food Habits
Although the collection period of scats for microhistological fecal analysis was
relatively short, the results clearly document the transition period from forbs to shrubs.
Lupinus spp. comprised the bulk of the relative density in August (64.89%) and
September (61.01%). By October, bitterbrush accounted for 52.20% of the relative
density. This collection was made during the cool, wet field season, so I would expect
the shift to shrubs to occur earlier in a normal, drier year.
Estimation of plant utilization at deer locations identified several species that were
not included in the fecal analysis. Utilization of plants was only calculated at deer
locations where no fresh sign of livestock was present to reduce the possibility of
confusing utilization by livestock with that of deer. Mackie (1970) identified most of the
same foods in central Montana for mule deer that I did in southcentral Idaho.
Multivariate Analysis
The results of the model-building process were informative although not
particularly useful. A relatively low predictive success (39.9%) was obtained with the
input of 3 labor-intensive variables: overhead cover and lateral cover (measures of
52
structure) and shrub cover (a measure of forage availability). Such site-specific variables
could not be acquired from most Geographic Information Systems (GIS) so they would
have to be measured in the field. The most important variable in the model, overhead
cover, was generally highest in the Riparian, Mountain Brush, and Aspen cover types.
These cover types comprised only 16.4% of the study area, yet 78.9% of all deer
locations were in one of these types. These cover types could easily be identified by a
GIS or by one of several methods of remote imaging. Equipped with only knowledge of
deer use of cover types and a GIS, a manager could predict deer locations at least as
accurately as the model.
/
. CONCLUSIONS
Mule deer does in the Bennett Hills were able to satisfy their basic needs in a
relatively small area indicating that the habitat quality in sites selected by the radiocollared does was quite high. The presence of the Mountain Brush cover type was a
major contributor to the high quality of the range, and its quality and quantity may be the
key indicator of habitat quality in the high elevation steppe of southcentral Idaho. This
cover type should be monitored and maintained at least at current levels. I identified
only 2 major threats to this cover type under current land uses: I) overstocking of cattle
could degrade stands, and 2) the tradition of burning large tracts of sage in the fall could
reduce acreage and canopy coverage of stands.
The short-term effects of a dry year on mule deer habitat use appeared to be
minimal. Deer shifted to cooler cover types and aspects. Overlap with cattle habitat use
may also have increased. An extended drought lasting several years or more would
probably magnify these effects.
Although the effects of livestock on mule deer could not be quantified in this study,
I believe the impact is minimal at current stocking rates. Cattle were only on the study
area for approximately the same period of time as deer, June through October. They
were moved off the area during winter due to high snow depths. During the summer/fall
period, deer and cattle dietary overlap is minimal and forage production is high.
54
I could detect little or no response of female mule deer to hunting seasons in the
Bennett Hills other than increased movement out of established summer/fall home
ranges. Hunting pressure is fairly high in this area with 800 to 2000 doe permits issued
for Unit 45 in 1993 and 1994. Hunting season, rather than snow depth or temperature,
may have triggered the fall migration to winter range. If this is true, early movement due
to hunting has the potential to prematurely deplete limited forage on winter range. Based
on normal weather patterns, deer could easily remain on the summer range for an
additional 3 to 6 weeks in the fall. A trial delay of the hunting season in Unit 45, with
continued aerial monitoring of radio-collared deer, may reveal if deer would utilize the
extra time on summer/fall range.
Management Implications
1. Mountain Brush and Aspen cover types should be maintained at current levels.
2. Burning of sagebrush tracts which contain significant amounts of bitterbrush
should be discouraged.
3. Riparian condition should be monitored and steps taken if degradation is
detected.
:
4. Consider the possible effect of hunting season on the timing of the deer
migration.
REFERENCES CITED
56
Austin, D. D., and P. J. Umess. 1985. Values of four communities for mule deer on
ranges with limited summer habitat. J. Range Manage. 38:167-171.
Bailey, J. A. 1984. Principles of wildlife management. John Wiley and Sons, New
York, N. Y. 373pp.
Bodurtha, T. S., J. M. Peek, and J. L. Lauer. 1989. Mule deer habitat use related to
succession in a bunchgrass community. J. Wildl. Manage. 53:314-319.
Brown, C., andM. McDonald. 1988. Mule deer ecology. Project W-160-R-15 job
progress report, Idaho Dept, of Fish and Game. 61pp.
Burt, W. H. 1943. Territoriality and home range concepts as applied to mammals. J.
Mammal. 24:346-352.
Carson, R. G., and J. M. Peek. 1987. Mule deer habitat selection patterns in northcentral Washington. J. Wildl. Manage. 51:46-51.
Case, C. W. 1981. Soil survey of Camas County area, Idaho. U. S. Dept, of Agriculture,
Soil Conservation Service, Washington, D. C. 139pp.
Currie, P. O., D. W. Reichert, J. C. Malechek, and 0. C. Wallmo. 1977. Forage
selection comparisons for mule deer and cattle under managed ponderosa pine. J.
RangeManage. 30:352-356.
Dickinson, T. G., and G. W. Gamer. 1979. Hbme range and movements of desert mule
deer in southwestern Texas. Proc Annu. Conf Southeast. Assoc. Fish and Wildl.
Agencies. 32:267-278.
Eberhardt, L. E., E. E. Hanson, and L. L. Cadwell. 1984. Movement and activity
patterns of mule deer in the sagebrush-steppe region. J. Mammal. 65:404-409.
Fielder, P. C., and E. McKay. 1984. Vegetation types used by mule deer fawns, midColumbia River,(Washington. Northwest Sci. 58:80-84.
Freedman, D., R. Pisani, and R. Pjurves. 1978. Statistics. W. W. Norton and Company,
Inc. New York, N. Y. 506pp.
Hayes, C. L., and P. R. Krausman. 1993. Nocturnal activity of female desert mule deer.
J. Wildl. Manage. 57:897-904.
Heezen, K. L., and J. R. Tester. 1967. Evaluation of radio telemetry data with special
reference to deer movements. J. Wildl. Manage. 31:124-141.
57
Hosmer, D. W., Jr. and S. Lemeshow. 1989. Applied logistic regression. John Wiley
and Sons, New York, N. Y. 307pp.
Jennrich, R. I., and F. B. Turner. 1969. Measurement of non-circular home-range J
Theor. Biol. 22:227-237. ,
'
Kie, J. G., J. A. Baldwin, and C. J. Evans. 1994. CALHOME: Home range analysis
program. Version 1.0. U. S. Forest Service, Fresno, Ca.
— , C. J. Evans, E. R. Loft, and J. W. Menke. 1991. Foraging behavior by mule deer:
the influence of cattle grazing. J. Wildl. Manage. 55:665-675.
Kufeld, R. C., D. C. Bowden, andD. L. Shrupp. 1988. Habitat selection and activity
patterns of female mule deer in the Front Range, Colorado. J. Range Manage
41:515-522.
Loft, E. R. 1988. Habitat and spatial relationships between mule deer and cattle in a
Sierra Nevada forest zone. Ph.D. Thesis, Univ. California, Davis. 144pp.
— , J. W. Menke, and J. G. Kie. 1991. Habitat shifts by mule deer: the influence of
cattle grazing. J. Wildl. Manage. 55:16-27.
Loomis, J. B., E. R. Loft, D. R. Updike, and J. G. Kie. 1991. Cattle-deer interactions in
the Sierra Nevada: a bioeconomic approach. J. Range Manage. 44:395-398.
Lund, R. E. 1993. MSUSTAT, statistical analysis package. Microcomputer Version
5.20. Montana State Univ., Bozeman.
Mackie, R. J. 1970. Range ecology and relations of mule deer, elk, and cattle in the
Missouri River Breaks, Montana Wildl. Monogr. 20pp.
— . 1978. Impacts of livestock grazing on wild ungulates. Trans. N. Amer. Wildl. Nat.
Resour. Conf. 43:462-476.
Marcum, C. L., and D. O. Loftsgaarden. 1980. A non-mapping technique for studying
habitat preferences. J. Wildl. Manage. 44:963-968.
Mohr, C. O. 1947. Table of equivalent populations of North American small mammals.
Am. Midi. Nat. 37:223-249.
National Climatic Data Center. Climatological data, 1993,1994. Idaho. U. S. Dept,
of Commerce, Asheville, N. C.
58
Noe, H. R. 1991. Soil survey of Elmore County Area, Idaho. U. S. Dept, of Agriculture,
Soil Conservation Service, Washington, D. C. 500pp.
Patric, E. F., T. P. Husband, G. G. McKiel, and W. M. Sullivan. 1988. Potential of
52^162^164^ WiIdIife research alonS coastal landscapes. J. Wildl. Manage.
Ragotzkie, K. E., and J. A. Bailey. 1991. Desert mule deer use of grazed and ungrazed
habitats. J. Range Manage. 44:487-490.
Rautenstrauch, K. R., and P. R. Krausman. 1989. Influence of water availability on
movements of desert mule deer. J. Mammal. 70:197-201.
Robinette, W. L., D. A. Jones, G. Rogers, and J. S. Gashwiler. 1957. Notes on tooth
development and wear for Rocky Mountain mule deer. J. Wildl. Manage. 21:134-
Rost, G. R., and J. A. Bailey. 1979. Distribution of mule deer and elk in relation to
roads. J. Wildl. Manage. 43:634-641.
Sargent, G. A., L. E. Eberhardt, and J. M. Peek. 1994. Thermoregulation by mule deer
in arid rangelands of southcentral Washington. J. Mammal. 75:536-545.
SAS Institute Inc. 1989. SAS/STAT user's guide, version 6, 4th Ed., Vol. 2. SAS
Institute Inc., Cary, N. C. 846pp.
Short, H. L. 1977. Food habitats of mule deer in a semidesert grass-shrub habitat. J.
Range Manage. 30:206-209.
Springer, J. T. 1979. Some sources of bias and sampling error in radio triangulation J
Wildl. Manage. 43:926-935.
Unsworth, J. W. 1992. Mule deer ecology. Study Plan for EDFG, Nampa, ID.
Uresk, D. W., and V. A. Uresk. 1982. Diets and habitats on mule deer in South-central
Washington. Northwest Sci. 56:138-147.
Wallace, M. C., and P. R. Krausman. 1987. Elk, mule deer, and cattle habitats in central
Arizona. J. Range Manage. 40:80-83.
West, N. E., and R. W. Wein. 1971. A plant phenological index technique. Bioscience
21:116-117.
59
Willms, W., A. McLean, R. Tucker, and R. Ritcey. 1979. Interactions between mule
deer and cattle on big sagebrush range in British Columbia. J. Range Manage
32:299-304.
Wood, A. K., R. J. Mackie, and K. L. Hamlin. 1989. Ecology of sympatric populations
of mule deer and white-tailed deer in a prairie environment. Mont. Dep. Fish,
Wildl. and Parks. 97pp.
Worton, B. J. 1989. Kernel methods for estimating the utilization distribution in home
range studies. Ecology 70:164-168.
Zar, J. H. 1984. Biostatistical analysis, second edition. Prentice-Hall, Inc., Englewood
Cliffs, N. J. 620pp.
60
APPENDIX
FREQUENCY, MEDIAN, AND MEAN COVER CLASS
OF VEGETATION IN COVER TYPES
Table 10. Frequency (percent of plots with plant species)/median/mean cover classes (1=0-1%, 2=>l-5%, 3=>5-25%,
4=>25-50%, 5=>50-75%, 6=>75-95%, 7=>95%) of vegetation recorded at deer locations, by cover type, on
________ the Bennett Hills study area, 1993 and 1994._________________
Species
Grass
Low Sage High Sage Sage/Bitter. MtnBrush Riparian
Aspen
N=
2
60
18
31
246
28
127
Trees
Populus tremuloides f<5m)
5/2Z2.3
6/1/1.5
48/2/2.4
7/3/3.0
71/2/2.6
Populus tremuloides f>5m)
3/2Z2.5
17/2/2.4
11/4/4.3
95/4/4.4
Salix spp.
22/4/3.9
96/5/5.3
11/3/3.0
Shrubs
Acer sp.
I/2/2.7
2/3/3.0
Amelanchier alnifolia
'28/1/1.6
55/2/1.8
48/2/1.8
41/2/2.4
4/1/1.0
43/2/2.3
Artemisia arbuscula
3/3/3.0
Artemisia cana
2/1/1.0
Artemisia longiloba
11/3/3.0
3/3/3.0
r
Artemisia ludoviciana
6/1/1.0
Artemisia tridentata
50/2/2.0 100/4/3.9 98/5/4.8 100/4/4.3 25/2/2.0
4/4/4.0
24/2/2.0
Berberis repens
50/1/1.0
28/1/1.8
18/1/1.5
3/2/2.0
76/2/2.3
28/2/1.9
Ceanothus sp.
50/1/1.0
6/2/2.0
g/3/2.8
43/3/3.0
2/1/1.5
Chrvsothamnus nauseosus
50/1/1.0
22/2/2.4 .42/3/2.5
28/2/2.2
6/2/1.6
4/3/3.0
ChrvsOthamnus viscidiflorus 100/3/3.0 83/2/2.4
88/2/2.4
71/2/2.3
10/2/1.7
19/2/1.8
Prunus spp.
50/2/2.0
20/2/1.7
28/2/2.2
42/1/1.6
98/5/4.5
48/3/2.9
Purshia tridentata
44/2/1.8
22/2/1.8 100/3/3.1
3/2/1.9
4/3/3.0
I/2/2.0
Ribes spp.
50/1/1.0
44/2/2.3
40/2/2.0
23/2/1.9
22/2/2.1
21/1/1.2
51/2/2.3
Rosa sp.
11/1/1.5
5/1/1.3
13/1/2.0
14/1/1.3
50/2/1.8
17/2/1.6
Svmphoricarpos sp.
50/2/2.0
44/2/2.6
70/2/2.3
48/2/2.1
91/3/2.6
25/2/2.4
95/3/2.8
Tetradvmia canescens
50/2/2.0
6/2/2.0
17/2/1.8
32/2/1.6
1/1/1.0
1/1/1.0
Forbs
Achillea millefolium
50/2/2.0
33/1/1.3
43/2/1.6
45/2/1.8
8/1/1.3
18/3/2.8
30/2/1.7
Agastache sp.
6/2/2.0
10/1/1.5
13/1/1.5
54/2/1.7
65/2/2.0
Table 10. Continued.
Species
Allium spp.
Astragalus filipes
Balsamorhiza sagittata
Calochortus sp.
Castilleja appleeatei
Castilleja sp. (other)
Cirsium spp.
Crepis spp.
Crvptantha sp.
Eqisetum sp.
Eribgonum spp.
Geranium sp.
Helianthella sp.
Lithosperma sp.
Lupinus leucophvllus
Lupinus spp. (other)
Mentha sp.
Penstomen spp.
Potentilla spp.
Solidago spp.
Sohaeralcea erossulariifolia
Tregopodon spp.
Urtica sp.
Wvethia sp.
Low Sage High Sage Sage/Bitter. MtnBrush
6/2/2.0
10/2/1.7
33/3/2.8
7/3/2.8
1/2/2.0
35/2/1.8
100/2/2.0 67/3/2.7
45/2/2.1
3/1/1.4
8/1/1.2
6/2/2.0
1/1/1.0
6/1/1.5
6/1/1
2/1/1.O
2/1/1.3
* 28/2/2.0
7/2/1.8
10/2/2.0
50/1/1.0
6/1/1.0
22/1/1.1
13/1/1.3
12/1/1.0
6/1/1.O
17/1/1.4
6/1/1.5 •
28/2/2.4
7/2/2.0
13/2/2.5
2/2Z2.5
Grass
61/3/2.5
6/1/1.O
50/1/1.0
50/3/3.0
17/1/1.3
56/2/2.4
22/1/1.5
50/1/1.0
75/2/2.4
17/1/1.0
2/2/2.0
32/1/1.2
5/2/1.7
70/2/2.5
10/1/1.0
2/1/1.O
3/1/1.5
3/1/1.O
3/1/1.O
2/2/2.0
2/1/1.O
39/2/2.4
10/1/1.0
3/2/2.0
26/1/1.1
-
2/2/1.8
2/1/1.O
4/2/2.O
12/1/1.2 .
Riparian
Asnen
4/2/2.0
8/1/1.3
4/1/1.O
4/2/1.6
25/1/1.0
10/1/1.0
6/1/1.6
2/2/2.5
4/3/3.O
6/2/1.6
20/1/1.5
I/2/2.0
9/1/1.2
...................
42/2/1.9 . 2/1/1.8
10/1/1.0
11/1/1.4
2/1/1.O
9/2/1.8
4/1/1.I
1/1/1.0
1/3/2.7
50/2/2.1
7/1/1.O
11/2/1.7
18/2/1.6
. 13/2/2.0
2/2/2.0
30/1/1.4
5/1/1.5
9/2/2.4
3/1/1.0
3/1/1.O
2/2/2.0
Table 11. Frequency (percent of plots with species)/median/mean of cover classes (1=0-1%, 2=>l-5%, 3=>5-25%,
4=>25-50%, 5=>50-75%, 6=>75-95%, 7=>95%) of vegetation, by cover type, as recorded at random
________ locations in the Bennett Hills, Idaho, 1993 and 1994.____________________________________
Species
Grass
Low Sage High Sage Sage/Bitter. MtnBrush Riparian
Aspen
N=
30_______ 55_______ 140
25_______ 35_______ 6
8
Trees
Populus tremuloides f<5m)
4/3/3.O
3/3Z2.5
37/3/2.8
17/2/2.0
75/3/3.2
Populus tremuloides f>5m)
1/2/2.5
4/3/3.0
9/1/1.7
33/3/3.0
88/4/4.0
Salix spp.
3/2/2.0
1/2/2.5
17/3/3.3 100/4/4.2 25/3/3.5
Shrubs
Acer sp.
2/3/3.O
17/1/1.0
13/1/1.0
Amelanchier alnifolia
7/1/1.5
22/1/1.5
36/1/1.5
52/2/1.9
77/2/2.1
33/2/2.5
50/2/2.5
Artemisia arbuscula
2/2/2.0
13/2/2.0
Artemisia cana
2/1/1.O
Artemisia longiloba
7/1/1.O
35/3/2.7
12/3/3.0
17/1/1.0
13/1/1.0
3/3/3.O
Artemisia ludoviciana
3/3/3.0
Artemisia tridentata
50/2/2.1
76/4/3.5 100/4/4.4 100/4/4.0 66/3/3.3
50/3/3.0
63/3/3.0
3/3/3.0
9/2/1.8
Berberis repens
2/2/2.0
4/2/2.0
66/2/2.2
25/2/2.5
CeanOthus sp.
8/1/1.5
4/2Z2.2
49/3/2.6
25/2/2.5
Chrvsothamnus nauseosus
7/2/2.5
16/2/1.6
16/2/2.1
36/2/1.9
23/2/1.8
25/1/1.5
Chrvsothamnus viscidiflorus
33/2/1.8
49/2/2.1
71/2/1.8
60/1/1.5
54/2/1.9
33/1/1.5
38/2/1.7
3/3/3.O
5/1/1.3
16/2/1.8
Prunus spp.
16/1/1.8
91/3/3.6
17/2/2.0
63/3/2.8
Purshia tridentata
10/2/1.7
18/2/1.7
29/2/1.6 100/3/2.9 49/2/2.0
13/1/1.0
7/1/1.O
2/1/1.O
24/2/1.8
Ribes spp.
8/2/2.0
29/2/1.8
50/2/2.3
38/2/2.0
3/1/1.0
2/1/1.O
16/1/1.1
Rosa sp.
8/1/1.0
37/2/1.7
83/2/2.0
38/2/1.7
7/3/3.O -13/1/1.4
50/2/1.8
Svmphoricarpos sp.
36/2/2.6
97/2/2.2
50/3/2.7 100/3/2.8
7/1/1.5
9/1/1.3
Tetradvmia canescens
2/2/2.0
12/1/1.3
20/2/1.9
Forbs
Achillea millefolium
37/1/1.8
41/1/1.3
27/2/1.9
12/2/1.7
40/1/1.2
67/2/2.0
38/1/1.3
6/1/1.5
Aeastache sp.
2/2/2.0
31/2/1.8
33/2/2.0
63/2/2.2
Table 11. Continued.
Grass
3/2/2.0
3/4/4.0
13/2/2.0
10/1/1.7
7/1/1.O
17/1/1.2
3/1/1.O
7/2/2.5
37/2/2.0
7/1/1.O
3/1/1.O
3/2/2.0
4072/2.1
3/1/1.O .
3/1/1.O
Low Sage High Sage Sage/Bitter. MtnBrush
2/1/1.O
28/2/1.7
4/1/2.0
14/2/2.1
8/1/1.O
39/2/1.8
48/2/1.8'' 40/2/1.6
22/2/2.3
4/1/1.2
11/1/1.3
2/1/1.3
9/1/1.3
11/1/1.3
4/1/1.O
3/1/1.O
27/2/1.7
14/1/1.0
12/1/1.0
40/1/1.0
20/1/1.0
10/1/1.0
■
4/1/1.0
3/1/1.O
11/1/1.5
9/2Z2.2
2/1/1.0
67/1/1.5
60/2/1.6
65/2/2.1
43/1/1.2
26/1/1.1
4/1/1.O
23/1/1.1
4/2/2.5
5/1/1.6
12/2/1.7
26/2/1.8
30/1/1.2 . 16/1/1.3
43/2/1.7
11/1/1.2
6/1/1.4
3/1/1.O
8/1/1.O
62/2/1.9
40/2/1.7
23/1/1.0
55/2/2.3
9/1/1.2
5/1/1,O
16/1/1.1
6/1/1.3
2/1/1.6
1/1/1.0
2/1/1.0
8/1/1.5
13/3/2.8
2/3/3.O
1/171.0
O
.......
oo
Species
Allium spp.
Astragalus filipes
Balsamorhiza sagittata
Calochortus sp.
Castilleia appleeatei
Castilleja sp. fother)
Cirsium spp.
Crepis spp.
Crvptantha sp.
Eqisetum sp.
Eriogonum spp.
Geranium sp.
Helianthella sp.
Lithosperma sp.
LupinUs leucophvllus
Lupinus Spp. fother)
Mentha sp.
Penstomen spp.
Potentilla spp.
Solidago spp.
Sphaeralcea erossulariifolia
TregopOdon spp.
Urtica sp.
Wvethia sp.
11/1/1.5
20/1/1.3
9/2/1.7
17/1/1.0
3/1/1.O
Riparian
Aspen
33/2/2.0
25/2/2.0
13/2/2.0
13/1/1.0
17/1/1.0
17/2/2.0
17/1/1.0
33/1/1.0
25/1/1.5
25/1/1.0
33/1/1.0
13/1/1.0
33/1/1.5
33/2/2.0
25/1/1.0
38/2/2.3
33/1/1.5
17/3/3.0
17/1/1.0
25/3/3.0
13/1/1.0
MONTANA STATE UMVERSfTY LIBRARIES
3 1762 10248775 6
ijriCAzOMAHA
NE.